证监会应该封锁Anthropic保护公众的IPO
马特·舍尔/2026年11月2日
Matt Scherer是最近的报告《Brace for Interact：在AI泡沫暴发时保护人民和经济》的作者。
在AI行业的财务方面，事情很少象看起来那样。 AI生态系统的循环商业交易，金融工程的广泛使用，以及其它的杂乱无章的财政，使得投资者，监管者，或为退休而储蓄的工人很难获得AI公司的准确财务情况。
Anthropic在计划11月股市首映前的上星期
该公司的顶级金融人物看起来相当令人印象深刻，在2026年中，收入似乎迅速增长，并且有一个相当健康的资产负债表——表面上是这样的。
然而，挖得更深一点，就会出现更暗淡的景象。 公司最新财务披露的脚注显示，支出承诺将近9 000亿美元，其中大部分都延伸到20世纪30年代。 公司2026年的财务结果没有受到正式审计，这使得其不可靠，但第三季度的结果表明，公司的收入增长在夏季放缓至攀升。
Anthropic将不得不扭转这一趋势，维持非常的收入增长，以履行其巨大的义务。 但挤压Anthropic的金融压力——事实上整个AI产业——将使领先的模型开发商难以做到这一点。 Anthropic的账单到期后
Anthropic必须攀登的金融山，对于一个在股市上市的公司来说是前所未有的。 虽然Anthropic有令人印象深刻的现金储备960亿美元，但经过修改的注册表揭示至少5,820亿美元不可撤销的支出承诺只提供给5家公司(Amazon,AMD,Broadcom,Google,Microsoft)，以获得培训和运行Anthropic基因AI模型所需的巨大计算力。
尽管这些债务并不出现在Anthropic的资产负债表上，但从法律和实际意义上讲，这些债务实际上是债务：这些债务是Anthropic根据合同今后必须支付的款项，无论公司是否有能力支付。
此外，尚不清楚8,950亿美元合同承诺的其余3,000亿美元中，有多少最终将落到Anthropic的资产负债表上。 随着该公司继续谈判进一步的大规模基础设施和计算合同，Anthropic很可能在年底以1万亿美元或更多的合同承诺。
至少可以说，任何公司都很难获得履行这些义务所需的收入。 但AI模型开发者的边缘正受到数种力量的挤压，这将使该任务对Anthropic产生特别的挑战。
关键力量挤压AI产业的边缘
Anthropic的花费是否合理？多项研究表明，LLMs没有加快生产力，也没有为大多数企业带来积极的投资回报，包括来自那些在AI上表现优劣的组织的研究。 一项研究担心以企业为重点的Anthropic，发现在2025年夏天(46%)到2026年(33%)之间，在工作中使用AI的美国人数量大幅下降。
Anthropic的支出承诺表明，该公司相信它能够说服企业为规模越来越大、成本更高的模型付费。 但越来越多的迹象表明，大多数企业对支付更强大更昂贵的模型并不感兴趣。 当今年夏天Anthropic 发布其多音节的5号线的前沿型号时，其商业接收量微乎其微。
即使未来的Anthropic模型吸引了企业，公司也会遇到第二个障碍：AI行业的疯狂竞争。 Anthropic是一个拥挤的硅谷生态系统的一部分，许多井上公司(包括Google,Meta,Microsoft,OpenAI，和SpaceX，以及Anthropic)提供基因AI模型，在质量和能力上基本无法区分。 因此，当竞争对手发现显示甚至暗示有商业潜力的用途时，AI公司可以迅速提供竞争产品。
在这种环境下，公司很少能够负担价格上涨，因为这样做只会促使客户转而与竞争对手竞争。 急于抢占客户，领先于他们的IPO，导致Anthropic和OpenAI到发布，今年夏天的价格战仍在进行。 提供开放权重模型的中国公司的竞争对价格造成了进一步下行的压力，而能够在当地使用电话和膝上型计算机的模型越来越多也是一样——不需要昂贵的数据中心。
这导致了第三个问题：培训和运行AI模型所需的芯片和数据中心费用特别高，这给Anthropic的成本带来了强大的持续上升压力。 在芯片、服务器、联网设备、电力和冷却基础设施、建筑费用和其他开支之间，一个1千兆瓦(GW)数据中心估计需要310亿美元，一些部件的价格正在飞涨。
这是1GW。 Anthropic已承诺在未来几年内拥有15个GW的数据中心容量，公司估计最终需要20-25个GW，这解释了公司在支出承诺中花费了一半的10亿美元以上。
这仅仅是有形基础设施的成本。 一旦一个数据中心建立并运行，训练AI模型需要使用数以万计平行运行的芯片，这些芯片需要大量能量来发电，运行热，需要保持冷却。 此外，每次有人提示一个模型回答一个问题，创建一个图像，或执行一个任务时，都需要额外的计算功率，能量，以及冷却来处理查询，生成一个响应或完成任务。
最后，有一些外部因素拖延了数据中心的建设，使费用增加。 全国各地普遍和数据中心对AI的强烈反弹，每一次延迟都会限制Anthropic的计算能力，Anthropic所依赖的产生其急需的收入，而由于伊朗战争和供应链挑战而增加的费用可能会促使Anthropic的运营成本仍然更高。
Anthropic公司如果IPO成功，将进一步增加其现金储备，但其业务的残酷经济学不可能很快改善。 即使需求大幅回升，激烈的竞争仍将对价格施加无情的下行压力，即使高昂的成本和资源制约使其从相反的方向挤出。 这对Anthropic来说是一个致命的组合，因为其收入增长放缓，资产负债表外的债务庞大，是巨大的固定成本。
如果Anthropic现在公开，它会把额外的风险堆积到一个股票市场上，这个股票市场在历史上已经被高估，而且过度集中在技术部门少数紧密相连的巨头公司。 这已经造成了历史上破坏性的股票市场崩溃的风险；以前从未如此庞大的财富取决于少数公司的业绩，而它们的财富是密不可分的。 如果Anthropic达到(或更确切地说达到)金融悬崖，在公共交流中增加另一个巨大的、无利可图的技术巨头将会加剧这两个因素，并造成更大的传染风险。
这还意味着，可能具有毁灭性的AI坠机事件所造成的后果中，越来越多的将落在普通人身上。 作为一家私营公司，Anthropic的股份绝大多数归公司内部人士、技术巨头和风险投资家所有。 一旦公司公开发行，股票开始流入401(k)s、退休账户和指数基金，当挤压Anthropic的利润开始被咬时，就会有工人、退休人员和零售投资者越来越多地被搁置。
证券交易委员会有权阻止这种情况发生。 如果该机构认为公司的注册声明 " 表面上是不完整或不准确的 "，它可以推迟或阻止公司公开。 考虑到该公司的预测书只包括了声称最大收入增长期间未经审计的财务结果，将它最重要的未来义务置于其登记说明的脚注中，并继续宣布新的11位数的财务承诺作为其IPO的临近。
证监会应封杀Anthropic的股票市场，直到公司连续提供两个季度的已审计财务报表，披露公司支出承诺的所有重要信息。 不这样做会对工人、投资者，并最终对金融制度和经济造成严重后果。
本文由 AI图鉴自动抓取并翻译，版权归原作者所有；如需查看原文请点下方链接。
The SEC Should Block Anthropic’s IPO to Protect the Public
Matt Scherer / Nov 2, 2026
Matt Scherer is co-author of the recent report Brace for Impact: Protecting people and the economy when the AI bubble bursts.
Things are seldom what they seem when it comes to the finances of the AI industry. The AI ecosystem’s circular business dealings, extensive use of financial engineering, and otherwise murky finances make it hard for investors, regulators, or workers saving for retirement to get an accurate financial picture of AI companies.
That is particularly true of Anthropic, which filed the registration statement for its initial public offering (IPO) last week ahead of a planned November stock market debut.
The company’s topline financial figures look quite impressive for a startup, with revenues that appear to have grown quickly over the course of 2026 and a fairly healthy balance sheet — on the surface, that is.
Dig a little deeper, however, and a far bleaker picture emerges. Footnotes to the company’s most recent financial disclosures revealed nearly $900 billion in spending commitments, most of them stretching into the 2030s. The company’s financial results for 2026 have not been subjected to formal auditing, which makes them less-than-reliable, but the third-quarter results suggest that the company’s revenue growth slowed to a crawl over the summer.
Anthropic will have to reverse that trend and sustain extraordinary revenue growth to meet its vast obligations. But the financial pressures that are squeezing Anthropic — and, indeed, the AI industry as a whole — will make it exceptionally hard for the leading model developer to do so. If the company goes public, workers, retail investors, and retirees will increasingly be left holding the bag when Anthropic’s bills come due.
The financial mountain that Anthropic must climb is unprecedented for a company making its stock market debut. While Anthropic has an impressive $96 billion in cash reserves, its amended registration statement reveals at least $582 billion in non-cancelable spending commitments to just five companies (Amazon, AMD, Broadcom, Google, and Microsoft) to gain access to the vast computational power required to train and run Anthropic’s generative AI models.
Even though they do not appear on Anthropic’s balance sheet, those obligations are effectively debts in a legal and practical sense: they are payments that Anthropic is contractually required to make in the future, regardless of whether the company can afford to do so.
Moreover, it is not clear how much of the remaining $300 billion or so of the $895 billion in contractual commitments will ultimately land on Anthropic’s balance sheet. With the company continuing to negotiate further massive infrastructure and compute contracts, Anthropic may well end the year with a trillion dollars or more in contractual commitments.
Generating the revenues necessary to meet those obligations would be, to say the least, quite difficult for any company. But the margins of AI model developers are being squeezed by several forces that will make that task extraordinarily challenging for Anthropic.
Key forces squeezing the AI industry’s margins
First, it isn’t clear that there is nearly enough demand for generative AI products and services to justify Anthropic’s spending. Multiple studies, including from organizations notable for their bullishness on AI, suggest LLMs are not accelerating productivity or otherwise generating positive returns on investment for most businesses. Worryingly for enterprise-focused Anthropic, one study found that the number of Americans using AI at work fell significantly between the summers of 2025 (46%) and 2026 (33%).
Anthropic’s spending commitments suggest that the company believes it will be able to convince businesses to pay for ever-larger and more expensive models. But there are growing signs that most businesses are not interested in paying for more powerful and expensive models. When Anthropic released its much-hyped Fable 5 frontier model this summer, its commercial reception was tepid.
Even if a future Anthropic model captivates businesses, the company would run into a second obstacle: the AI industry’s frenzied competition. Anthropic is part of a crowded Silicon Valley ecosystem where numerous well-heeled corporations (including Google, Meta, Microsoft, OpenAI, and SpaceX, along with Anthropic) offer generative AI models that are largely indistinguishable in their quality and capabilities. As a result, AI companies can quickly offer competing products whenever a rival identifies a use case that shows even a hint of commercial potential.
In such an environment, companies can rarely afford to raise prices because doing so will prompt customers simply to switch to a rival. The rush to grab customers ahead of their IPOs led Anthropic and OpenAI to launch a price war this summer that remains ongoing. Competition from Chinese companies offering open-weights models puts further downward pressure on prices, as does the increasing availability of models that can run locally on phones and laptops — no expensive data centers required.
That leads to the third problem: the extraordinarily high costs of the chips and data centers needed to train and run AI models, which put powerful and continuous upward pressure on Anthropic’s costs. Between the chips, servers, networking equipment, electricity and cooling infrastructure, construction costs, and other expenses, a 1 gigawatt (GW) data center costs an estimated $31 billion, and the prices for some components are soaring.
That is for 1 GW. Anthropic has committed to 15 GW of data center capacity in the coming years and the company estimates it will ultimately need 20-25 GW, which explains the company’s half-trillion-plus dollars in spending commitments.
And that is just the cost of the physical infrastructure. Once a data center is up and running, training AI models entails using tens of thousands of chips running in parallel that require massive amounts of energy to power, run hot, and need to be kept cool. Moreover, each time someone prompts a model to answer a question, create an image, or perform a task, additional computational power, energy, and cooling are needed to process the query and generate a response or complete the task.
Finally, there are external factors that are delaying data center construction and piling on additional costs. The fierce and still-rising backlash to AI in general and data centers in particular across the country has scuttled a number of data center projects and delayed many others. Every delay limits the computational capacity Anthropic is counting on to generate the revenue it will desperately need, and rising costs due to the Iran War and supply-chain challenges could drive Anthropic’s operational costs still higher.
Anthropic will further pad its cash reserves if its IPO is successful, but the brutal economics of its business are not likely to improve anytime soon. Even if demand picks up significantly, intense competition will apply unrelenting downward pressure on prices even as high costs and resource constraints squeeze it from the opposite direction. That is a deadly mix for Anthropic in light of its slowing revenue growth and massive off-balance-sheet debts, a vast and immovable fixed cost.
If Anthropic were to go public now, it would pile additional risk onto a stock market that is already historically overvalued and disproportionately concentrated in a handful of tightly interconnected mega-cap companies from the tech sector. That has already created a risk of a historically damaging stock market crash; never before has such a huge amount of wealth depended on the performance of a small number of companies whose fortunes are inextricably linked. Adding another huge, unprofitable tech giant to public exchanges would exacerbate both of those factors and create a higher risk of contagion if (or, rather, when) Anthropic reaches a financial precipice.
It also would mean that an ever-larger share of the fallout from a potentially devastating AI crash would fall on ordinary people. As a private company, Anthropic’s shares are overwhelmingly owned by corporate insiders, tech giants, and venture capitalists. Once the company goes public and shares start getting funneled into 401(k)s, retirement accounts, and index funds, it will be workers, retirees, and retail investors that are increasingly left holding the bag when the forces squeezing Anthropic’s margins start to bite.
The Securities and Exchange Commission (SEC) has the power to stop that from happening. The agency can delay or stop companies from going public if it believes a company’s registration statement is “on its face incomplete or inaccurate in any material respect.” That legislative language might as well have been written with Anthropic in mind, considering that the company’s prospectus includes only unaudited financial results for the period where it claims the greatest revenue growth, relegates its most significant future obligations to the footnotes of its registration statement, and has continued to announce new 11-digit financial commitments as its IPO approaches.
The SEC should block Anthropic’s stock market debut until the company provides two consecutive quarters of audited financial statements, with all material information about the company’s spending commitments disclosed. Failing to do so could have dire consequences for workers, investors, and ultimately the financial system and economy.
Auto-fetched and translated by AI Atlas. All rights belong to the original author.